Title: Dynamics, structure, and organization of complex lipid membranes

Abstract

The plasma membrane is a highly dynamic soft interface, made up of lipids, proteins, and carbohydrates, which constantly assemble and disassemble complex structures across multiple length scales, to accommodate the ever-changing needs of the cell. At the shortest length scale, the relatively high diffusivity of individual biomolecules enables rapid bimolecular interactions, such as receptor-ligand binding. However, biological membranes also host higher-order domains of proteins and lipids, such as T-cell receptor clusters, which can reach tens of microns in size, and thus have relatively low diffusivity. Using non-equilibrium transport via the cytoskeleton, cells can overcome the kinetic limitations to assembling and transporting such large surface domains. The ability to dynamically remodel surfaces at the microscale to control chemical reactions is highly desirable for interfacial materials and may endow the next generation of materials with improved chemical selectivity, the ability to self-heal, among other properties.

In this thesis I approach the challenge of studying dynamic, non-equilibrium materials by simultaneously studying the mechanisms by which cells organize their surfaces, while also developing simplified model interfacial materials to study interactions between lipid membranes and actin. I first present novel molecular probes that insert into the plasma membrane of living cells in culture and bind soluble antibodies to probe the local osmotic pressure of the glycocalyx brush. I show that this crowded brush is highly heterogeneous in all directions: x, y, and z, with most crowding being localized within a few nanometers of the surface, and outside of lipid/protein clusters and domains. I then present model membranes that phase-separate in 2D, with filamentous actin selectively adsorbed only to the liquid-disordered phase. The actin avoids the liquid-ordered dispersed phase, while causing the viscous liquid-ordered droplets to deform in response to its elasticity. The lipid domains stiffen the actin network, while the actin changes the domain ripening behavior from that of a purely viscous system. I further perturb the system using mechanical activity, delivered by myosin motors, to dynamically control these behaviors. I find that actomyosin activity generates 2D flows in the membrane, which decreases the domain relaxation time by 23x. Under certain flow fields, it also accelerates the domain coarsening rate by a factor of 2. These studies of lipid membranes fall into a broad category of work on droplet coarsening under shear and in elastic networks, but are the first to address these questions experimentally with a quasi-2D viscoelastic material.